h a l f b a k e r y
Oh yeah? Well, eureka too.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
This Idea starts with a modified version of an ordinary
deadbolt, which has a keyhole on the outside and a
on the inside. It also has a little mechanical switch on
inside, and an internal gearing arrangement.
Suppose that from the outside, you turn the key
the deadbolt from the locked to the unlocked
position. This switch lets you decide whether or not it
be clockwise or anti-clockwise, that moves the
from the locked to the unlocked position. (In practice,
you will probably need to move the deadbolt to the
half-way-extended position, before being able to flip this
Now imagine 10 such deadbolts along the edge of the
All take the same key, but some you set the deadbolt
switches for clockwise locking, and others you set for
anti-clockwise locking. The door has to have a
plate (see link) that covers the whole edge, so a thief
cannot see which deadbolts are latched.
So, after setting the deadbolt switches, you go outside
lock, say, 5 of the deadbolts. If the burglar has a copy
your key, but doesn't know the code, then efforts to
the correct deadbolts will lead to locking other deadbolts.
Eventually, the thief will give up. No burglar ever
wants to be seen taking a long time to open a door.
After the thief has given up and gone away, and you
return home, you might think at first you have a
because now the code of which deadbolts to unlock has
But, the REAL code is the pattern of 10 switch-settings!
You know which way to turn each of the 10 deadbolts to
move them to the unlocked position. So you unlock the
door, go inside, and you can now change the pattern
the deadbolt switches.
As mentioned in the main text. This plate needs to be long enough to cover all the deadbolt latches, so the burglar can't see whether a particular deadbolt is locked or unlocked. [Vernon, Mar 21 2014]
||You could probably defeat this by simply pulling on the door
as you turn the key in each lock. One deadbolt will bind on
the door jamb, and when you find and successfully unlock
it the door will move a tiny bit, and that deadbolt will
resist being locked again (since it will now bind against the
jamb). This is much the same way that tumbler locks are
||[ytk], yes, that is a good point. I suppose in the
never-ending conflict between security makers and
security breakers, the next step would be to have a
sensor connected the door, to detect such a "pull",
and to disable all the locks. Let the key rotate
endlessly in any of them, with nothing happening. :)
||There's also the issue that for most deadbolts, lock
and unlock feel and sound different.
||So basically you turn your door into a combination lock with 2^10=1024 combinations?
||I'm not convinced that's enough combinations to protect a house.
||[MechE], in my experience that mostly has to do
with how tight-fitting the deadbolt goes into the
notch in the door-frame. We might prefer a loose
fit here, relying on the door-handle-latch to be a
||[Loris], yes, it is not a huge number of
combinations. However, trying each one takes
time, and like I wrote, the thief doesn't want to
spend a lot of time on unlocking the door. Note I
assumed the thief had the help of a valid key --
most thieves don't have that. When you add the
factor that you can easily change the combination,
a thief that tries some now, and then some more
later, and then some more later-still, could in
theory be kept stymied, provided you actually do
change the combination reasonably often (once a
||Let it SMS you every time there's an error. If you accidentally make a mistake, you have to put up with the small annoyance of paying to be reminded of the fact by your phone; if not, maybe there's time enough to deal with the intrusion. (But then similar comments to the combination lock apply -- why not just get an alarm. Also simplicity is lost.)
||Maybe you could encode some kind of delay in
between the trials in some fashion before another
attempt can be made.
||If this was an idea to use existing hardware in a clever new way to increase security or if it was a convienient or efficient way to provide two factor authentication, I might see the benefit. Unfortunately, this idea required a new reversible deadbolt design, and you end up with a system that is annoying to operate and is fairly vulnerable to being watched through a telescope. It may also be hard to ensure that there isn't some way to get a wire behind the security plate to feel out all the deadbolt positions.
||This seems like a cumbersome way of making a
combination lock with 1024 settings. It also solves a
non-existent problem: very, very few burglaries are
committed by picking locks, having copies of keys or
whatever. It's generally faster and more reliable to
gain entry another way.
||Are you-all forgetting this is the HalfBakery, where
Ideas are sometimes only semi-sensible?
||Sometimes the point of increasing security isn't to deter break-ins but to cure paranoia. I think having ten switchable deadbolts does that quite well.
||As [scad] points out though, the code can be stolen merely by watching the owner lock the door.
||How about, instead of mounting the locks separately along the edge of the door, you mount them all in a stack behind the same keyhole. (Ten might be difficult, maybe three would suffice.) In this way, the code would be entered using one long-shafted key which has to be slid in and out to set each lock position. You would still need separate mortices to prevent feeling the solution.
||In fact, why limit ourselves to two possible states for each lock? The locks could be continuously variable with only a certain key angle corresponding to an open lock. The key could sit only in the first mechanism, with the second mechanism engaged by a tab on the back of the first, and so on, so that entering the code would be like twisting a combination padlock, a series of specific angle alternating direction turns. This would probably baffle even those in possession of both the key and the code.
||I notice that nobody else has mentioned the obvious flaw in
this plan, namely that it can be defeated by a 12-lb
sledgehammer and thirty seconds without adult
supervision. This makes it no more secure than any
||I mention this not because I am a luddite moron who has
missed the point (I ain't and I hasn't), but because the
overwhelming majority of professional burglars do not
bother picking locks. In locales where 'possession of
burglar's tools' is a felony add-on, the basic definition of
said usually reads 'a crowbar and a short-handled hammer
in any sort of bag or sack'. The only people who bother with
picking locks are locksmiths and people who have legally
obtained a finely-crafted set of lock picks and taught
themselves how to pick locks simply because they enjoy
knowing that they know how to pick locks.
||I never found out how the visitors next door got in. They must've broken in very quietly, because there was no time to run or phone. I wonder if our professional robbers now pick locks instead of the old fashioned lo-tech brick through a window, then hide in the vicinity to see if anyone noticed. One thing I do know is that waiting for people to leave the house is now very "old school". When the very latest in cutting edge crime technique reaches you (and it's only a matter of time, I'm afraid), you'll find that the true pro waits till you're *at* home, not away from it.
||I may have mentioned in the past that I dislike the attitude of "don't bother, burglars don't pick locks."
||For the same reason that some people enjoy learning to pick locks, others enjoy trying to make them unpickable.
||Good point. I'm always forgetting to check my pragmatism
at the door when I come here.